[gmx-users] Re: Re：the mdp parameters for localpressure calculation using gromacs-4.0.2_localpressure

[Jianguo Li]

________________________________
From: KONG Xian <xianshine at gmail.com>
To: ljggmx at yahoo.com.sg
Cc: gmx-users at gromacs.org
Sent: Sunday, 21 August 2011 19:20:50
Subject: Re：the mdp parameters for localpressure calculation using 
gromacs-4.0.2_localpressure

 
Thanks for your kindly reply. 
I used the LINCS calculated the pressure field. The distinct between protein and 
the lipid bilayer is obvious from the pressure field. 

But the lipid bilayer and the water is not very obvious , they seems to be a 
uniform from the pressure field calculated. As in Ollila’s original paper: 3D 
Pressure Field in Lipid Membranes and Membrane-Protein Complexes, there is an 
distinct separation region between membrane and the water.
Maybe the LINCS is the problem.
Another problem is perhaps the localpressure software is for CG model? My model 
is all-atom model and the simulation time length is not long enough, maybe this 
is the cause of the results? 

-- From my limited experience, the local pressure calculation converges  very 
poorly. It takes more than 50 ns of production run in my case  to get a 
converged profile. You can divide the production run into several  
sub-trajectories to judge the convergence.

KONG Xian
2011/8/21
 
Hi, Kong Xian,
 
It is better to use SHAKE instead of LINCS since LINCS does not directly yield 
pairwise forces, as shown in the paper: Erik Lindahl and Olle Edholm. Spatial 
and energetic-entropic decomposition of surface tension in lipid bilayers from 
molecular dynamics simulations. JOURNAL OF CHEMICAL PHYSICS, VOLUME 113, NUMBER 
9, 3882b3893.
 
best regards,
Jianguo 
 
 
 
 
________________________________
From: KONG Xian <xianshine at gmail.com>
To: gmx-users at gromacs.org
Sent: Friday, 19 August 2011 23:50:01
Subject: [gmx-users] the mdp parameters for localpressure calculation using 
gromacs-4.0.2_localpressure
 
I have sent this email days ago, but I got no answer.
Hope someone would saw it this time. 
Sorry for disturbing.
Dear all:
         I am using gromacs-4.0.2_localpressure to calculate the local pressure 
of my system.
         I have a question.
         When rerun the mdrun from gromacs-4.0.2_localpressure I used a new .mdp 
file. There are some changes of the new .mdp file according the original one. I 
have 2 questions:
1.      I changed the The coulomb interaction type from PME to reaction field 
with epsilon_r=1 epsilon_rf=78;  Is this change feasible?
2.      I used LINCS for all bonds in the simulation, and I still use the LINCS 
for all bonds when calculate the local pressure. Is this way right? 
 
Thanks for any reply.
Best wishes,
KONG Xian
Tsinghua, Beijing, China
 
 

__________ Information from ESET NOD32 Antivirus, version of virus signature 
database 6397 (20110821) __________

The message was checked by ESET NOD32 Antivirus.

http://www.eset.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20110822/522cdec9/attachment.html>